With a certification engineering hat on, which is my background, I'd probably take a somewhat softer line than Bob on this one ..
CAO 20.7.4.1.a suggests (my bolding) ..
An aeroplane must not take off at a weight in excess of the least of the weights determined in accordance with subparagraphs (a) to (d):
(a) a weight at which the take-off distance required under subsection 6 for the pressure height, temperature, runway slope (if in excess of 1%) and wind component along the runway, is equal to or less than the take-off distance available in the direction of take-off. Approved declared conditions may be used instead of actual pressure height and temperature;
while CAO 20.7.5.1.a suggests (my bolding) ...
Except in an emergency, an aeroplane must not land at a weight in excess of the least of the weights determined in accordance with subparagraphs (a) and (b):
(a) a weight at which the landing distance required in accordance with subsection 10 for the pressure height, temperature, runway slope (if in excess of 1%), and wind component along the runway at the time of landing, is equal to or less than the landing distance available in the direction of landing. Approved declared conditions may be used instead of actual pressure height and temperature;
First point is that the CAO 20.7 series dates back a long ways and is the operational restatement of what were the earlier certification words in the various (then) ANO 101 series documents - now, post Ron Yates report, unceremoniously binned WPB. I think we need to cut Gavin some slack as the certification side of things is not his (or any of his operations colleagues - with the exception of folk such as Miles Gore-Brown who originally comes from the certification engineering side of the house) - stamping ground.
Adding, just as speculation, the concession possibly dates back to the early days of strip length certifications, commonly used for some of the old rag-bag aeroplanes. In this manner, those aircraft would have a concession for nominally level surface operations to preclude the need for flight test work to develop any hard data for the CofA.
Second, if one reads the words carefully, the "must" requirement applies if the slope exceeds 1%. That is to say, the words are silent on what one should do for smaller slopes.
My suggestion is that this is a concession, acknowledging that the effect of a shallow slope is not overly critical and recognises that the declaration of slope is a tad airy-fairy in that the slope is rarely uniform along a runway centreline. Certainly, when I was actively involved in airline performance scheduling work, I took account of the actual slope variations along the runway centreline when assessing just how I was going to go about accounting for slope in the sums and the final RTOW data generation.
For light aircraft there is no benefit to be gained, in reality, by ignoring the shallower slopes - ie, one, sensibly and pragmatically, ought to use the slope in the calculations. If the slope is less than 1 %, you can choose not to use it .. but why would you do so ?
At the end of the day, if the calculated distance for a light aircraft is critically limiting, acknowledging the approximations and such which go into the usual data generation, one really should consider putting a bit more pad into the weight limit on the day .. I know I always did on lighties and never came to grief .. or even a significant fright so far as runway length was concerned.
Looking at the P-charts, which were developed by DCA (a precursor of CASA), the equations which went into the charts were fairly simple. The overall thrust was to end up with charts which were a bit conservative but not to the extent that the Industry would discount them as being silly. Certainly, my experience in generating a number of the P-charts followed this philosophy. Furthermore, having had a close working relationship with the DCA performance certification side of the house, I certainly can confirm that their motivation was strongly aligned with the philosophical concepts suggested above.
As to the OEM charts, well, I would need to do some flight test checks to form an opinion for any given data ....
